A Study on Possible Solutions to the Challenges Associated with

Limited Survey Locations in Community Noise Measurement Based

on Noise Mapping in China

Jiping Zhang a), Heng Ma b), Weike Wang c), Zheming Wang d)
Institute of Environmental Policies & Standards
Zhejiang Research and Design Institute of Environmental Protection
109 Tian Mu Shan Road
Hangzhou, 310007, China

ABSTRACT
The spacing grid method (SGM) is one of several survey methods used for 2D ground
surface urban area community noise measurement (CNM). In order to solve the problems
caused by limited time, availability of funding and an excessively large number of
measurement locations, it is necessary to decrease the total number of sampling locations
for the CNM urban sound functional area grids to a minimum using mathematical
optimization. The proper grid spacing recommended by the relative guideline is 500 m ×
500 m to 1 km × 1 km. This paper will discuss and illustrate the practicability of this
treatment. However, it is important that the locations number for every unit area should be
considered. Theoretical simulations show that if the location density is too small, the SGM
related to the total location number may result in failure because the noise in urban areas
can usually propagate to 200–300 m at most, considering barriers and the reflector
effectiveness of different obstacles. In addition, this paper proposes to use popular noise
mapping instead of SGM. The noise map not only provides information on urban noise
distribution in 3D space considering insulation and reflections from buildings, but it also
helps to reduce the total survey locations efficiently. The reduced measurement results can
then be used for the calibration of the noise mapping.

1 INTRODUCTION

Since 1980, the standards for environmental noise in China have improved considerably.
Further advancement in this area should focus on the optimization of methods, and improved
accuracy and efficiency based on China’s actual situation, to enhance the tools required for

a)
email: jpzhang_daniel92@163.com
b)
email: 47959702@qq.com
c)
email: 30457184@qq.com
d)
email: 747533658@qq.com
sound environmental quality protection and noise pollution mitigation or control. Community
noise measurements (CNMs) to determine existing conditions are important. CNMs are usually
conducted during both daytime and nighttime and are used for two major aspects of noise
management. The first are cases related to environmental impact assessment (EIA) and
completed project acceptance (CPA) for engineering construction projects and zoning planning
of urban sound functional areas. The second are cases related to urban area sound environmental
management (UASEM).
The spacing grid method (SGM) is one of several survey methods used for urban area
community noise monitoring (CNM) on the ground surface in 2D space. The objective of such
an approach is to observe, analyze, and assess the annual change and tendency of the total sound
environment in a city. However, in order to solve the problems associated with limited time,
availability of funds and an excessively large number of measurement locations, the total
sampling locations for CNM of the urban sound functional area grids should be minimized by
exploiting mathematical optimization. The recommended grid spacing based on the relative
guideline can is 500 m × 500 m to 1 km × 1 km. This paper will discuss and illustrate the
practicability of this treatment, however the location number for every unit area should be
considered. Theoretical simulations show that if the location density is too small, SGM related to
the total location number may result in failure. This is because noise in urban areas can usually
propagate to 200–300 m at most considering barrier and reflectors effectiveness from different
obstacles. This paper proposes to use a popular noise mapping instead of SGM. The noise map
not only provides information on urban noise space distribution in 3D by considering the
insulation and reflections of buildings, but it also helps to reduce the total number of survey
locations efficiently. The reduced measurement results can then be used to calibrate the noise
mapping.

2 2D SGM AND 3D NOISE-MAPING IN URBAN AREA SOUND ENVIRONMENTAL

MANAGEMENT (UASEM)

2.1 Historical Cases of 2D Community Noise Survey by SGM

An early case of grid noise measurement survey 1 was referenced in Medford, Massachusetts,
during March, April, and May 1971, to assess the importance of transportation noise in a typical
urban community. This was done to obtain validation data for a mathematical model developed
to simulate such noise levels and to learn how best to measure community noise levels so that
more efficient sampling techniques could be used in later surveys elsewhere. Noise
measurements were performed at 49 locations that were distributed uniformly across the
community. The noise levels were influenced primarily by highway traffic, with a major
highway arterial corridor producing the highest noise levels in the community. Scheduling and
funding consideration dictated that no more than three months could be allotted to the planned
survey which meant that measurements could reasonably be taken at approximately 50 locations
throughout the city allowing for probable weather and technical delays. To establish complete
and uniform coverage, a rectilinear grid network was superimposed over the map of the city with
a north-south, east-west orientation at 2200-ft spacing. These grid lines are shown in Fig. 1.
 A map of Medford,
Massachusetts, and the
locations of the 49
measurement sites

Fig. 1 –illustrating the major transportation tracks through the city, the grid overlay to allow uniform
coverage for noise measurements.
Measurements were then taken at each grid intersection, insofar as possible, to ensure
uniform coverage. In addition, microphone locations were selected as close to the grid
intersections as practical, depending on the physical topography, accessibility, and the desire to
avoid abnormal local noise sources. Each measurement took approximately 40–60 min. Main
measurements were made only during the morning rush hour and at noon. Each measurement
was analyzed statistically, and the important statistical descriptors were reported.
Another good case 2 of similar sampling strategy which was recently completed is from the
city of Boston. To identify potential monitoring sites, the city was divided into 500 m × 500 m
grid cells using ArcGIS (Redlands, CA, USA). Sites inaccessible to monitoring, i.e., sites located
in the middle of a body of water or in the middle of a busy road, were discarded. Once a list of
all potential accessible sites was established (n = 650), 450 site locations were randomly selected
as shown in Fig. 2. Each site was monitored once and site location monitoring varied by time of
day (day: 7 am–7 pm; night: 7 pm–7 am) day of the week (weekday: Monday–Friday; weekend:
Saturday and Sunday) and season (Winter, Spring, Summer, Fall). The results indicate that low-
frequency noise is the dominant frequency component in urban environments. Urban
characteristics such as time of day, land use, proximity to major roads, bus/train routes, hospitals,
and neighborhood building density had a statistically significant influence on the median sound
level across the entire frequency spectrum. Community perception (ascertained via The Greater
Boston Neighborhood Noise Survey) maps suggest that areas closest to dense transportation
networks are not only the loudest but also are perceived to be the loudest by community residents.

Legend:
the points in red are the monitoring sites

Fig. 2 – A Map of Noise Monitoring Site Locations in the City of Boston

2.2 Urban Area Community Noise Monitoring by 2D SGM in China

In China, UASEM begun in the 1980’s. For instance, Zhejiang Provincial Environmental
Monitoring Center organized the first large-scale urban noise investigations in three cities of
Hangzhou, Ningbo, and Wenzhou from 1981 to 1982. The total effective locations of the area
environmental noise grids reached as high as 410. At the end of 2010, the total grids locations in
the eleven cities in Zhejiang Province were 1953 3. Up to 2017, this value 4 was 2060 and
covered an area of 1285.44 km2. The average equivalent sound level weighted with the area is
55.2 dBA. The established sophisticated sound monitoring grids for sound environmental quality
have succeeded in facilitating reports of sound monitoring results to the China National
Environmental Protection Monitoring Center. These reports highlight the existing sound quality
and their changing tendency to provincial environmental protection agencies that support the
management of all levels of government with scientific databases, in addition to providing the
basis for technical assistance for inspections and assessments.
For the 0–3rd sound environmental functional area (SEFA) in China, monitoring
requirements for general GSM 5~7 include dividing the areas awaiting SEFA into many equated
large squares, for example, 1000 m × 1000 m. For unconnected built-up areas, the grids can be
unconnected. If the grids are located in an area that is 100% water surface or is unable to be
monitored such as a forbidden zone, or in an area where more than 50% is un-built area, then
they are ineffective. The grids formed should cover the total area of the GSM, and the total
number of effective grids should be more than 100. Every monitoring point should be located at
the center of every grid, where it satisfies a general outdoor condition. When the center locations
are such that they are not suitable for measurements such as on water surface, in forbidden areas,
and the lanes of traffic road, the monitoring position should be relocated to a position as close to
the grid center as possible. The monitoring locations should meet the general outdoor selection
requirements according to the standard GB3096 in China. Specifically, the height of the sensor at
the measurement position should be 1.2–4.0 m. Each monitoring session should be performed in
the work time of the daytime and the nighttime between 22:00–24:00. It can be postponed when
there is insufficient time. The monitoring day should avoid holidays and other abnormal working
days. For the aforementioned duration, each monitoring session at a location should measure
equivalent sound levels Li lasting 10–20 min at the center location of all the grids to calculate the
arithmetic mean and standard deviation in Eqn. (1) and to assess quality according to Table 1,
and note the main source of the noise.

(1)

where, is the urban area daytime average equivalent sound level ( ), or the urban
area nighttime average equivalent sound level ( n)，both in dBA； is the measured
equivalent sound level at the center of the i th grid, in dBA; n is the total number of all
effective grids.

Table 1 – Class level of urban area sound environmental quality, in dBA

Class level First Second Th i r d Four F i ft h

Daytime average
≤50.0 50.1–55.0 55.1–60.0 60.1–65.0 >65.0
equivalent sound level ( d )
Nighttime average
≤40.0 40.1–45.0 45.1–50.0 50.1–55.0 >55.0
equivalent sound level( n)
Quality level Better Good General Bad Worse
 Based on the noise values at each grid center and their grids areas, the percentage area of
different noise impact levels, their criterion reach in daytime and nighttime, and the impacted
populations can be counted and estimated. The average value represents the total community
noise level in the sound environmental functional area.

2.3 2D Grids Requirement According to ISO and WHO

With the requirements of microphone positions relative to reflecting surfaces, noise levels are
normally calculated in grid points for use in noise mapping. In special cases where
measurements are performed, the density of the selected grid points 8 in an area depends on the
spatial resolution required for the study and the spatial variation of the sound pressure levels of
the noise. This variation is strongest in the vicinity of sources and large obstacles. The density of
grid points should therefore be higher in these places. In general, the difference in sound pressure
levels between adjacent grid points should not be greater than 5 dB. Intermediate grid points
should be added if significantly higher differences are encountered.
Various degrees of accuracy are required depending on the purpose of the calculation 9. The
necessary density of grid points used as a basis for mapping the noise levels in an area depends
on the purpose of the mapping. Noise-level variation is strongest in the vicinity of sources and
large obstacles. Therefore, the density of the grid points should be higher in such places. In
general, mapping of the difference in sound pressure levels between adjacent grid points should
not be larger than 5 dB for overall noise exposure. When selecting noise-mitigation measures in
the form of either noise control hardware or economic compensation, the grid-point density
should be chosen so that the variation between adjacent points does not exceed 2 dB.

2.4 3D Noise Mapping in UASEM

The main international agreement and definitions for noise mapping 10~11 were established in
relation to the Environmental noise directive of the European Parliament and Council (Directive
2002/49/EC of 25 June 2002, commonly referred to as the END). The EU Member States are
required to produce strategic noise maps in their main cities, near main transport infrastructures
and industrial sites. The main goals of the END are to create a diagnosis of noise pollution in
Europe that can lead to action plans and noise management that can be implemented in terms of
action plans and acoustical planning. The term 'strategic' is very important in this definition
because the management of environmental noise must be for the long-term in a full area.
Using either approach, a grid of receivers must be defined in order to measure or calculate
noise levels. When results are obtained, using GIS tools, spatial interpolation must be applied in
order to obtain a continuous graphical representation of the sound levels. The END 5 dBA ranges
were used for the contour representation. The maps may be useful during planning stages, for
prior evaluation of action plans, or the determination of the most polluted areas. It is possible to
represent an evaluation of the number of people exposed to a specific dBA range using a
strategic noise map. Facade sound levels must be calculated or estimated from the previous map.
Simulation tools are very useful especially in the planning stages, where measurements are
not possible. Consultants can evaluate the effectiveness of action plans to guide appropriate
decisions. Examples of noise mapping software include LimA, CadnA, Code_Tympan, dB
Foresight, IMMI, Predictor, Olive Tree Lab Terrain, SoundPlan, noise3D online.
3 SOLVING PROBLEMS PRODUCED BY TOO-FEW LOCATIONS IN CNMS BY
VERIFICATION WITH BORROWED CASES

3.1 A Point Noise Case Proof with Theory in Room Acoustics

(1)Sampling element
We borrowed the virtual image concept in room acoustics to demonstrate the density of
SGMs (unit size of grid per area) is important in case of failure when the grid area is too large.
In Fig. 3, the assumed sampling elements (the researched grids of the urban areas) are
equivalently established with one represented noise source inside a basic large “Room (Box)”
and its virtual images. The four virtual walls and top ceiling of the virtual box are smooth or hard
with “0” for absorption or “1” for completive reflecting characteristic. The ground of the room is
formed with wood and soft earth for which the absorption can often be set as 0.512. The grid size
or the length and width of the box is assumed as a × a. The height of the box H should cover the
protected receiving points, such as 1–7 storey general multi-buildings, 8–12 storey high-
buildings, and 13–24 storey high-rise buildings. Here we recommend a height of 36 m.

        Legends:
 --Sound
source;
 --sound
receiving
        point;
        the room size:
aGaGH,
that is the grid
 size equals to
        aGa, and
        H is
recommended
as a height 36
m for a
        represent 12
        storey high
building

Fig.3 - Establishing a grid system with virtual images for a noise source in a virtual acoustical box

(2)Grid size of sampling interval

The sound level at the grid center can be estimated with the classical room sound prediction
models in Eqns. (2) – (4),

(2)
(3)

(4)

where, Lp is the sound level at the receiving point which is at the center of grid in this case and is
measured in dBA; Lw is the sound power lever of the noise source in the grid in dBA; r is the
distance between the noise source and the grid center, in meters; Q is the directivity of the noise
source where 1 is used if the source is suspending in space and 2 if the source is on the surface;
R is the room constant which represents the noise from the other grids; 0 is the ground feature
recommended as 0.5; a is the length or width of the grid in m; H is the height of the box which is
recommended as 36 m for a representative 12 storey high-building.
(3)Results and Analysis
Eqns. (2)–(4) simulate that the noise level at the grid center is from two components, one
from the noise source within the monitoring grid, and the other from the noise sources virtually
located in the grids. If the two parts are equal, we can derive Eqn. (5) from Eqns. (2)–(4),

(5)

When r < rm, the noise value at the researched grid is mainly from the noise source in this grid,
and when r > rm, the noise value will be mainly from the noise sources outside the researched
grid. This concept is parallel to the critical distance defined in room acoustics.
Based on Eqn. (5), we can determine the maximum size of the grid a × a by setting

Lp(Q/4r2)-Lp(4/R)≦3dB, (6)

So , (7)

let Q = 2, H = 36, , r≦a/ , and the calculate result is then a≦87 m.

Where, the distance r is between the two points when the monitoring or receiving point is located
at the center of one grid, and the noise source is near the center, and will not exceed the
boundary of the room. When the noise source is located at the center of the grid, the monitoring
point can be located around it without exceeding the boundary of the room so as to avoid the two
points occupying the same space.
The simulated result shows that when there is one noise source in each grid, the grid size
should be smaller than 87 × 87 m2 in the case of CNMs for UASEM. This size can satisfy the 5
dB contour requirement according to ISO or WHO relative standards. So the general sizes of 500
× 500–1k × 1k m2 should be too looser to cover such urban areas, considering the impact from
the noise sources in other grids. However, when the size of 87 × 87 m2 is adopted, the working
burden will be simultaneously increased.

3.2 Line Noise Source Case of Roads Critical Distances of Propagation Characteristics
Including Pre-existing Ambient Noise by SGM and Noise Mapping

(1)Sampling element
In Ref.13, based on the 2D grids requirement according to ISO and WHO and 3D
mathematical modeling by means of proprietary software using typical input traffic data, several
road situations involving noise attenuation with distance were assessed and discussed. The zone
of road critical distances is not only due to the road concerns but also to the preexisting ambient
environmental noise level. The main results are listed in Table 2. Apart from noise mapping, if
only the SGM is used to obtain road information within one grid of 500 m × 500 m–1 km × 1 km,
the density would be too low to include them.
 Table 2 – Critical distances for the case studies in Ref. 13
L Night background Noise barrier wall Critical
Cases Road type Average Annual Traffic，v/h
Noise level dBA or building distance, m
H.1 Expressway 38374pcu/d, Ratio Day/night=85% 0 / 43.4 / 43.4 - / - / Wall 348 / 385 /215
H.2 Expressway Night510v/h, Elevated, Surface 469v/h 0 / 42.1 / 42.1 - / - / Wall 642 / 686 /584
H.3 Class 1 highway Day:602v/h, Night:173v/h 0 / 43.8 / 43.8 - / - / Wall 159 / 186 / 101
H.4 Class 1 highway Night178 v/h 0 / 38.4 / 38.4 - / - / Wall 106 / 116 / 15
H.5 Class 2 highway 7956 pcu/d 0 / 47.1 -/- 141 / 221
H.6 Class 3 highway 2543pcu/d, Night: 42v/h 0 / 41.9 -/- 34 / 39
U.1 Urban freeway Night:584 v/h 0 / 42.5 / 42.5 - / - / Row of buildings 222 / 254 / 30
Night:Freeway715v/h, 100%P,Arterial:
U.2 Elevated Freeway 402 0 / 46.6 / 46.6 -/- /Wall&Row of buildings Both 235 / 340 / 30
U.3 Arterial road Night:710v/h 0 / 48.3 / 48.3 - / - / Row of buildings 251 / 346 /30
U.4 Collector road 17653pcu, Night: 331v/h 0 / 41.6 / 41.6 - / - / Row of buildings 102 / 159 / 14
U.5 Local road 1036pcu/d, Night:7 v/h 0 / 49.1 -/- 6 / 27
U.6 Local road 1036pcu/d, Night:7 v/h 0 / 46.3 -/- 5 / 11

(2)Grid size of sampling interval

It is shown in Table 2 that the noise level of urban busy arterial roads can generally
propagate and impact locations as far as approximately 300 m away, including the contribution
of local background noise. When noise barriers or the first row of buildings are introduced to the
roads, the additional benefits are that the critical distances are narrowed. Besides, highest noise
levels from highways or expressways can reach more than 600 m away, but these traffic flows
are rare in urban areas.
(3)Results and Analysis
Even in the case of the 300 m buffer distance, it is still shorter than the aforementioned 500
× 500–1k × 1k m2 grid size because of the features of traffic noise propagation distance, while
road noise is the general source in the researched grids. With regard to the insulation of noise
barriers or the first row of buildings, the local sound fields will be too complex for the simple
SGM approach. Moreover, only the size of noise mapping can satisfy the principle of 5 dB
contour requirement according to ISO or WHO relative standards.
Compared with the assessment of air pollution, the atmospheric impact can reach very
distant places with winds in complex air movement, so the environmental statistical methods of
SGM are often applied and is generally more dependable than calculations or simulations. Noise
impact can only reach places locally and have relationships with noise source distributions and
reductions, or reflections from buildings and obstacles. Thus the methods in air EIA may not be
suitable for community noise assessment.
The size of noise mapping can easily satisfy the 5 dB contour requirement according to ISO
or WHO relative standards compared to that of SGM. If the SGM can be applied to a distance of
300 m within a local area, a lower density may fail to show detailed information. Inversely, a
higher density of SGM may introduce a heavy burden on work and cost. Noise mapping is only
suitable, practical, and dependable in UASEM.

3.3 A Borrowed Case of Community Noise Survey in a Built-up Area by SGM

(1)Sampling element
An excellent example of this situation in reference14 was borrowed to demonstrate our idea
in Fig.3, where a built-up area with 13000 inhabitants for which a noise map and a noise
reduction plan is produced for a total area of approximately 2.3 km2. A heavily traveled road
with a mean daily traffic density of about 15000 vehicles per 24 h leading through a built-up area
has been calculated in a 10 m grid.
 (2)Grid size of the sampling interval
There are two types of sampling elements. In the first case, the grid size of the sampling
interval is 1 km × 1 km, and it is 500 m × 500 m in the other. For every element, it is also
separated into daytime and night. The noise levels at the grids centers were examined from the
noise mapping with the simulated values instead of monitoring them on site.
(3)Results and Analysis
Both the noise mapping and the grids results at their centers are shown in Fig. 4 and Table 3.

Case G-1: Case G-1:

Ld = 55.4 dBA Ln = 46.7 dBA
The case
area is
divided into
1 grids.
Grid size of
sampling
interval
=1 km × 1
km

Case G-1: Day time Ld=55.4 dBA Case G-1: Night time Ln=46.7dBA

Case G 4-2 Case G 4-2

Case G 4-1 Case G 4-1
Ld= 73.9 dBA Ln = 63.9 dBA
Ld= 66.2 dBA Ln = 61.6 dBA

The case
area is > 35.0 dB > 35.0 dB

divided into >

>
40.0 dB
45.0 dB
>
>
40.0 dB
45.0 dB
> 50.0 dB
4 grids. >
>
50.0 dB
55.0 dB > 55.0 dB
> 60.0 dB > 60.0 dB
Grid size of >
>
65.0 dB
70.0 dB
>
>
65.0 dB
70.0 dB
sampling Case G 4-1
>
>
75.0 dB
80.0 dB
Case G 4-1
>
>
75.0 dB
80.0 dB
Ln = 33.1 dBA
interval Ld= 41.9 dBA > 85.0 dB > 85.0 dB

=500 m × Case G 4-4

Case G 4-4
500 m Ld= 46.7 dBA Ln = 38.0 dBA

Day time in dBA: Average: 57.2 Nighttime in dBA: Average: 49.2

Case G4-1: Ld=66.2; Case G4-2: Ld=73.9; Case G4-1: Ln=61.6; Case G4-2: Ln=63.9
Case G4-3: Ld=41.9; Case G4-4: Ld=46.7 Case G4-3: Ln=33.1; Case G4-4: Ln=38.0

Legends:
Ground plan: 3D-view: cavalier:

Fig. 4 – Borrowed case: A built-up area with 13000 inhabitants for which a noise map and a
noise reduction plan was produced. The total area is approximately 2.3 km2.

Table 3 – Grid 1 km × 1 km verses the Grids 500 m × 500 m, in dBA

Grid size of
Case Main noise sources Daytime Ld Nighttime Ln
sampling interval
1 km × 1 km G-1 Road and build-up living Ld=55.4 46.7
G4-1 Road 66.2 61.6
G4-2 Road 73.9 Average: 63.9 Average:
500 m × 500 m
G4-3 build-up living 41.9 57.2 33.1 49.2
G4-4 build-up living 46.7 38.0
Noise build-up living Assessment
-1.8 -2.5
between 1 km × 1 km and 500 m × 500 m
 From Table 3, we can see the average sound levels of the area in a detailed sampling grid
size of 500 m × 500 m versus 1 km × 1 km. The case with G4-1–4-4 is much more precise than
the case of G-1. This implies that the later part has an error of 1.8–2.5 dBA compared to the
former.
Further, considering the 500 m × 500 m grid only, G4-1 and G4-2 grid or G4-3 and G4-4
have similar results because of the similar terrain and noise source distribution. In this situation,
the grid can be simplified by using a bigger grid. Otherwise, the method can result in failure such
as the case of G-1.
However, with noise mapping, a 10 m grid can be easily realized instead of SGM in with
higher dependability and accuracy compare to SGM.

4 CONCLUSIONS AND DISCUSSIONS

4.1 Conclusions

This paper suggested an approach to address the problem produced by too few survey
locations in community noise measurement with noise mapping and decreased survey locations.
From the highlighted cases involving room acoustics and the two simulations presented in
this paper, it can be seen that the density of grid spacing is very important in CNM for an entire
city. This is particularly in the case of insufficient survey locations which may result in the loss
of physical meaning of the monitoring approach. One of the reasons is that the heavy roads can
impact the local environment as far as 300 m away, so that a grid spacing of more than 500 m
may result in failure. Moreover, because of reflections and barriers caused by buildings and
obstacles, the sound fields may dramatically concentrate or extend in local urban areas. This can
cause difficulty in choosing the field CNMs locations on site. Even if monitoring locations are
recommended, they may not represent real practical situations. While it has been shown via
theoretical and experimental studies that the location of a SGM can be optimized and decreased
in some UASEM cases, this is based on the precondition of a uniform surface distributions of
outdoor sound sources and the avoidance of acoustical reflections and barrier reductions on site.
Under these circumstances, the density of the SGM can satisfy the 5 dB contour requirement
according to ISO or WHO relative standards.
Therefore, we propose that is appropriate to use popular noise map instead of SGM. The noise
map not only indicates urban noise spatial distribution in 3D by considering reflections and the
insulation of buildings and other obstacles, but it also helps to reduce the original total number of
survey locations efficiently. The reduced number of measurements can be used to calibrate the
noise map. Noise mapping method will modernize and update the tools of the UASEM of China
to facilitate greater accuracy of operation. The cost savings for SGMs can then be reinvested into
other devices and software for noise mapping.

4.2 Discussions

Firstly, before noise mapping became a mature approach and gained general acceptance,
SGMs were widely used in China for more than 30 years as an effective approach for reducing
workload, cost and time, in addition to playing an important role in UASEM. The results of such
studies have been included in annual environmental quality reports of numerous governmental
agencies at all levels. At the same time, the SGMs have been operated by local environmental
monitoring centers or stations with stable financial support from local governments. However,
their limitations which include the problem of insufficient survey locations are an increasing
concern. It is therefore urgent that China is encouraged to prioritize nation-wide noise mapping
using UASEM as a substitute SGMs by implementing trial usage, and development in cities such
as Beijing, Shanghai, Guangzhou, Suzhou, and Hangzhou, et al. This noise mapping method can
contribute much more than just a decrease in survey locations, by analyzing complex sound
fields that are not possible using only SGM.
Secondly, the meaning of urban community noise assessment in Eqn. (1) is only an average
index for assessment, for example, when comparing the noise levels among many cities, with
few relationships with the practical noise that impact the people living in a city or in determining
which standard criterion should be used. Urban noise mapping can reveal noise levels locally in a
detailed way in order to aid in noise control plan or precaution planning when the values exceed
the criterion set in the community noise standard.
Thirdly, noise mapping is also a helpful tool in solving problems15 related to short sampling
times in community noise measurements.

5 ACKNOWLEDGMENTS

This paper is the work of one research entitled A Cases Study on the Effectiveness of the
support technology, policies and regulations for environmental noise management in Zhejiang
Province of China, original from a research program No.2012F10037. Material presented in the
present paper is only a reflection of the authors’ academic views. It does not constitute in any
way in an official statement formulated in their affiliation of work.

6 REFERENCES

1. J. E. Wesler, “Community noise survey of Medford, Massachusetts”, J. Acoust. Soc. Am., 54,
985-995, (1973).

2. Erica Walker, “Characterization of Community Noise and Sound Levels in an Urban City:
Preliminary Results”, Proceedings of NOISE-CON2016, Providence, RI, 2016 June13-15.

3. Editing Committee, Volume of Environmental Protection, General Society of Zhejiang,

Vol.(8), Zhejiang People’s Press, Hangzhou, (will be published at the end of 2018). (in
Chinese)

4. Zhejiang Environmental Protection Agency, 2017 Annual Working Summary of

Environmental Noise Pollution and Control, March, 2018. (in Chinese)

5. GB3096-2008, Environmental quality standard for noise, (the latest version in 2008
including the measurement method; 1st version in 1993, separated 1st measurement version
in 1994 coded GB/T3222-94). (in Chinese)

6. GB/T15190-2014, Technical Specifications for Regionalizing Environmental Noise Function

(the latest version in 2014; 1st version in 1994 coded GB/T15190-1994,. (in Chinese)

7. HJ640-2012, Technical Specifications for environmental noise monitoring—Routine

monitoring for urban environmental noise. (in Chinese)
8. ISO/FDIS1996-1:2003 & 1996-2:2006, IDT. Acoustics － Description, measurement and
assessment of environmental noise, Part1: Basic quantities and assessment procedures,
(2003); Part 2:Determination of environmental noise levels, (2006). International Standard
Organization, Geneva, Switzerland.

9. Edited by: Bergitta Berglund, Thomas Lindvall, and Dietrich H. Schwela, Guidelines for
Community Noise, WHO, Ministry of the Environment, Institute of Environmental
Epidemiology, Geneva, Switzerland, Available as a pdf file over the Internet at:
http://www.who.int/docstore/peh/noise/Comnoise-1.pdf （1999）.

10. https://en.wikipedia.org/wiki/Noise_map.

11. Cadna A. State-of-art noise prediction software, Reference Manual. Release 4.4; DataKustic:
60~68.

12. L.F. Cohn and G.R. Mcvoy, Environmental Analysis of Transportation Systems [M], A
Wiley-Intersciencs Publication, (1982).

13. Jiping Zhang, MarcBuret, Shuoxian Wu, et al. Buffer distances for surface roads and
elevated highways correlated with pre-existing ambient noise. Proceedings of Internoise-
2014/ the 43th International Congress and Exposition on Noise Control Engineering,
November 16-19, 2014 in Melbourne; Congress Program, Session of Motor vehicle noise -
policy and regulation, D8b: paper IDNo.812.

14. DataKustik, Examples\05_Immissions\Option XL\XL_01.cna, 15.1 Noise Maps and their

Evaluation, Chapter 15 – Options X & XL, Reference Manual of Cadna A, Release 4.4:
page 15-1~82.

15. Jiping Zhang, Weike Wang, Heng Ma, Zheming Wang, Solving problems due to too-short
sampling times in community noise measurements: Remedies from case studies,
Proceedings of ICSV25-2018, would be on 8-12 July 2018, Hiroshima, Japan: Paper 310.